ABSTRACT
Two aspects of experimental design that are paramount to the success of a binary response bioassay with one explanatory variable are dose selection and sample size: These two factors affect the precision of estimation. A computer program, OptiDose, was especially designed to provide special guidance in the placement of doses and sample sizes that are necessary for precise estimation of response levels other than 50" or 90". OptiDose provides general guidance concerning the number of doses to use in a bioassay, but it was not designed specifically to study the number that would be theoretically optimal. At least two dose-fixing experiments are usually required to determine doses for a specialized bioassay. The first is the same type done for the general bioassay: a logarithmic series of dilutions must be done to tentatively identify the effective range for 10"–95" mortality. A well-designed experiment includes adequate sample sizes, complete replication, and doses carefully placed to estimate the LD or concentration of interest.
